Russell Stover's innovative Eskimo Pie treats were the smash-hit of 1920, and represented the culmination of long and careful experimentation with different techniques for adding a chocolate coating to ice-cream.
But when Stover was ready to patent his invention, he ignored his lawyer's advice to keep the patent narrow and relevant to his actual process, preferring to greedily seek a patent on any chocolate/ice-cream bar. As the Eskimo Pie grew in popularity, new companies invented new ways of combining chocolate and ice-cream and Stover sued them all.
Soon, Stover's legal bills exceeded even the prodigious revenues from the sales of his bars, and eventually he had to sell the company at fire-sale prices, starting over and founding his eponymous candy company.
As Charles Duan (who wrote the brilliant short story Stop the Music for us) writes, this is a parable for today's epidemic low-quality software patents that are overbroad to the point of meaninglessness.
If the Eskimo Pie Corporation had obtained a patent of more reasonable breadth, maybe it would have continued to innovate, dreaming up new treats to delight a willing public, rather than resting on the laurels of an overbroad patent. For patents can encourage inventions, but they can also forestall new ideas—follow-on innovations—that build upon those inventions.
Indeed, it is follow-on innovation that gave us the ice cream novelty most familiar to us today. When Christian Nelson first made his original I-Scream Bar, he put it on a stick for holding. Russell Stover convinced Nelson to take the stick out, saying that the stick made the product look too much like a lollipop. The Eskimo Pie ended up being sold stickless. The mass marketing of a chocolate-coated ice cream bar on a stick would require a new visionary who would see beyond the Eskimo Pie, a man named Harry Burt, who would go on to found the Good Humor company.
Ice Cream Patent Headache